The Agent Skills Directory

[SAFE]: The skill consists solely of Markdown documentation and a JSON evaluation file. No executable scripts (.sh, .py, .js) or automated command executions were found within the provided files.
[SAFE]: No obfuscation, prompt injection attempts, or persistence mechanisms were detected in the instructions or metadata.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest context from external project files (e.g., .agents/product-marketing.md). While this constitutes an attack surface for indirect prompt injection, the skill itself does not provide any high-privilege capabilities (like file writing or network operations) that could be exploited through these ingestion points.
Ingestion points: SKILL.md (reads .agents/product-marketing.md, .claude/product-marketing.md, or product-marketing-context.md).
Boundary markers: Absent in the instruction text.
Capability inventory: None; the skill provides only textual advice and does not include scripts or subprocess calls.
Sanitization: Absent.
[SAFE]: All references to external files are local within the skill's own package (references/) and contain legitimate business logic and psychological frameworks.

onboarding